Sindarin Phonetic Development (Part 71)

Sindarin Phonetic Development (Part 71)

Forums

S. final [i], [u] generally vanished; [-S{ĭŭ}|-uCu|-Sī] > [-Sø|-uCu|-Sĭ]

It is well known that short final vowels vanished in Sindarin and Noldorin, but treating this a single set of sound changes presents problems for other phonetic developments, most notably i-intrusion. As suggested by Elaran in a private Discord conversation on 2018-08-25, i-intrusion is much easier to explain if we assume final i-loss occurred later, after the loss of other short vowels. There are also (potentially related) issues with the loss of final u (see below).

Assuming final i-loss occurred separately, most of the final i vowels would already have vanished as part of the process of i-intrusion. If there were any remaining final i, they would appear only after consonant clusters, which resisted i-intrusion. It is possible any final i after consonant clusters also vanished as part of i-intrusion, but they mostly did so without any further phonetic effects because the resulting consonant palatization could not penetrate the cluster. Therefore, it seems likely that the timing and mechanics of final i-loss were distinct from those of other final vowel losses, worth discussing separately.

The problem of bereth: As pointed by Bertrand Bellet in his Vowel Affection in Sindarin and Noldorin (VASN), there is one problematic Sindarin example where i seems to vanish after a single consonant without the usual phonetic effects of i-intrusion, in the noun bereth “queen” and the name Elbereth:

This noun had a similar etymology in the Noldorin of the 1930s:

As pointed out by Bertrand Bellet, the Noldorin phonetic developments are not problematic, because [ei] sometimes became [e] in unstressed final syllables in Noldorin, and the phonetic development could have been:

  • barathī > *berethi (fronting) > *bereith (intrusion) > N. bereth (ei-reduction).

This ei-reduction does not occur in Sindarin, however, making this word hard to explain as a derivative of barathī(e); the expected form through normal phonetic development would be **beraith. Elaran suggested a possible solution to this quandary in a Discord chat in November 2018. Elaran noted that Tolkien introduced a more detailed explanation of bereth in The Road Goes Ever On in the late 1960s:

... but bereth actually meant “spouse”, and was used of one who is “queen” as spouse of a king (RGEO/66).

This new explanation hints that Tolkien may have re-etymologized this word as a derivative of the root √BER “marry”, and thus it may not have involved i-affection at all. As further support for this theory, in the notes where the derivation ✶barathī(e) > S. bereth appears (PE17/23), Tolkien provided an alternate origin of the name Elbereth that mentioned bereth “queen as spouse of a king” but without mentioning the primitive form barathī(e). These notes late 1950s or early 1960s might be the point in time when Tolkien changed his mind about this word’s derivation. If so, the derivation barathī(e) > bereth, might simply be a (discarded) remnant of Noldorin ideas.

Final u-loss: If short i-loss occurred later than other vowels, it is possible the same was true of short u-loss, since both of these were high vowels (front and back respectively) and there were similar specializations in their phonetic development in Primitive Elvish (to -e and -o). Regardless, it is convenient for discussion to treat the Sindarin and Noldorin loss of short final u separately from other short vowels, since it too seems to have some special developments. In particular, at one point Tolkien considered u-intrusion or w-intrusion analogous to i-intrusion:

N.B. final -w (left after loss of vowels) in Sindarin was dropped after labials (-mw > mm anyway): after other consonants became ŭ or was intruded like y but without alt[ering] of the preceding vowel. So matwā [>] madw̯ > maud or madu. teswā “[?chip]” > teχwā > teχw̯ > tewch (PE17/148).

There is no clear evidence of w-intrusion outside this specific note, so I think w-intrusion itself was probably a transient idea. However, this note does hint that in some circumstances, short final u could survive (unlike short final i). In this particular example, the final u in madu is a result of final -w and thus is not a true survival of a primitive final vowel. There are other examples, however, where primitive final seems to survive in Sindarin and Noldorin:

  • ngurū > S. guru “death” (PE17/87).
  • ON. nguru > N. gûr, gurw?, where the “?” after gurw indicates Tolkien was unsure of the correct development (Ety/ÑGUR, EtyAC/ÑGUR); the form guru appears in another entry in the Etymologies (Ety/WAN).

Another u-survival might be a factor in the etymology of N. guruth:

  • ON. ngurtu [> *(n)gurthu] > guruth (Ety/ÑGUR), perhaps with inversion of the final syllable.

There are other examples where a final u was lost, however:

  • kherū > S. hir [hîr] “lord, master” (Let/282).
  • Kh. Felakgundu > felaggundu > S. Felagund “Cave-hewer” (PM/352).
  • ᴹ√MAN > N. mân “departed spirit” vs. ᴹQ. manu (Ety/MAN).
  • ᴹ√KUND-Ū > N. †cunn (Ety/KUNDŪ; EtyAC/KUNDŪ).

Perhaps where ever two short u appeared in sequential syllables, a final u might survive. This might be related to the mechanism that prevented [u] from become [o] in these circumstances. As indicated by Felagund and N. cunn “prince”, it seems the final u did not survive if it was separated form the other u by a consonant cluster.

Final i-survival: There are also a couple of examples where final i survives in Sindarin and Noldorin. Final i is much less common than final u, because final u was often the result of the vocalization of final w after another consonant. This explanation does not work for final i, however, since it well known that a final y ended up intruding into preceding syllable as part of the process of i-intrusion. The two main examples of final surviving [i] are:

  • S. serni “shingle, pebble bank” < ✶sarniye (VT42/11).
  • N. fili plural of N. fela “cave” < ON. phelga (Ety/PHÉLEG).

In the first example, the likely development is ✶sarniye > *sarnīe > *sernī(e) > serni, so a plausible explanation for the survival of the final i is that it was long rather than short at the time of final vowel losses. In the second example, the phonetic development might also have been: *phelgi > *filʒi > *filī > *fili, with [-Cɣi] > [-Cī]. If so, then it seems likely that in those rare cases where a final ī was long rather than short at this late stage of phonetic development, it simply shortened rather than vanishing.

Conceptual Developments: As discussed in the entry on how short final vowels vanished, survival of primitive final and was the rule rather than the exception in the Gnomish of the 1920s:

C. culu. brindi. urthu. gwilthi. Phonologically -u, -i only refer to [primitive] -ū, -ī (PE11/14).

Final vowel loss (both short and long) seems to have mostly become the rule by the Early Noldorin of the 1920s, but there are hints of possible u-survivals: ᴱN. bagru “wares” (PE13/138); ᴱN. nabru “booty” (PE13/150).

The general loss of final i and u (alongside rare survivals) seems to behave similarly in later Noldorin and Sindarin. For a more detailed discussion on complexities final i developments, see the entry on how final [i] intruded into preceding syllable.

Submitted by Lokyt Thu, 03/21/2019 - 21:29

In reply to by Lokyt

Alright, I gave it a little digging. Now, if I'm not mistaken, you just assumed that at some point -ĭı̯->-ī- and that it was before the final vanished (and before the i-fronting as well).

I fully agree about the process itself. And as for the timing, it's not impossible either - after all, the only term post quem we know about is AFAIK the a-affection (necessary to get N. minei<mineı̯a out of miniı̯a before it would become *minīa), and that's developement #52, while the i-fronting and the loss od final -ĕ/-ă are #61 and #66 respectively.

However, I find it more likely that the -ı̯-, separating syllabic nuclei of different quality, remained distinct until the final disappeared. Only then the now final -ĭı̯>; and as a matter of fact, it could be the same process as https://eldamo.org/content/words/word-70600889.html: in both cases a final non-syllabic high vowel is "consumed" by a preceding instance of the same vowel (be it syllabic or not).